A coordinate input apparatus is available. This coordinate input apparatus is used to input a coordinate point designated by a pointer (e.g., a dedicated input pen, finger, or the like) so as to control a connected computer or to write characters, graphics, and the like.
Conventionally, as a coordinate input apparatus of this type, touch panels of various methods have been proposed or become commercially available. These touch panels are prevalently used since they allow easy operations of a terminal such as a personal computer or the like on the screen without using any special tools and the like.
Various coordinate input methods such as a method using a resistive film or an ultrasonic wave, and the like are available. A coordinate input method using light is also available, as disclosed in U.S. Pat. No. 4,507,557. This U.S. Pat. No. 4,507,557 discloses the following arrangement. That is, a retroreflecting sheet is formed outside a coordinate input region. Illumination units for illuminating an object with light and light-receiving units for receiving light, which are arranged at the corners of the coordinate input region, are used to detect the angles between the light-receiving units and a shielding material such as a finger or the like that shields light in the coordinate input region. Based on the detection results, the position pointed by that shielding material is determined.
Also, Japanese Patent Laid-Open No. 2000-105671, 2001-142642, or the like discloses a coordinate input apparatus which includes a retroreflecting member formed around the coordinate input region, and detects the coordinate position of a portion (light-shielded portion) where retroreflected light is shielded.
In the apparatus disclosed in, for example, Japanese Patent Laid-Open No. 2000-105671, the peak of a light-shielded portion by a shielding material, which is received by the light-receiving unit, is detected by a waveform process arithmetic operation such as differentiation. With this process, the angle of the light-shielded portion with respect to the light-receiving unit is detected, and the coordinate position of the shielding material is calculated based on that detection result. Also, Japanese Patent Laid-Open No. 2001-142642 discloses an arrangement in which one end and the other end of a light-shielded portion are detected by comparison with a specific level pattern, and the center of these coordinate positions is detected.
Note that the method of detecting a light-shielded position and calculating the coordinate position like as disclosed in Japanese Patent Laid-Open Nos. 2000-105671 and 2001-142642 will be referred to as a light shielding method hereinafter.
Furthermore, in such coordinate input apparatus based on the light shielding method, especially, when the coordinate input region has a large size, a demand is arisen for an application that allows a plurality of operators to input simultaneously so as to attain a more convenient and efficient meeting or the like. For this purpose, a coordinate input apparatus that supports a plurality of simultaneous inputs has been proposed.
In order to simultaneously input a plurality of coordinate positions, Japanese Patent Laid-Open No. 2002-055770 or 2003-303046, or Patent Registration No. 2896183 discloses a technique for detecting the angles of a plurality of light-shielded portions by one light-receiving sensor, calculating several input coordinate candidates from a combination of the angles of each sensor, and detecting an actually input coordinate position from these input coordinate candidates.
For example, when two coordinate points are input, a maximum of four coordinate points are calculated as input coordinate candidates, and actually input two coordinate points are determined and output from these four points. That is, in this determination, actual input coordinate points and false input coordinate points are selected from a plurality of input coordinate candidates, thus determining final input coordinate points. This determination will be referred to as true/false determination hereinafter.
As a practical method of this true/false determination, Japanese Patent Laid-Open No. 2003-303046 or Japanese Registration Patent No. 2896183 discloses the following technique. That is, first and second sensors are arranged on the two ends of one side of a conventional coordinate input region, so as to be spaced apart by a distance large enough to precisely calculate a coordinate position pointed within the coordinate input region. Furthermore, a third sensor is arranged at a position between the first and second sensors, so as to also be spaced apart from the first and second sensors by a distance large enough to precisely calculate a coordinate position pointed within the coordinate input region. On the basis of angle information of the third sensor, which is different from those of the first and second sensors, the true/false determination is made for a plurality of pieces of angle information detected by the first and second sensors.
As a detection method of a light-shielded portion shielded by a pointer such as a finger or the like by a light-receiving element which forms a light-receiving unit in the conventional coordinate input apparatus of the light-shielding method, Japanese Patent Laid-Open No. 2001-142642 has proposed a technique that detects the center between one end and the other end of a light-shielded portion in an output signal of the light-receiving element as the detection coordinates corresponding to the pointer. Simply, Japanese Patent Laid-Open No. 2002-055770 has proposed a technique that detects the position of a light-shielded portion as a pixel number corresponding to the light-shielded portion detected by the light-receiving element. These detection methods of a light-shielded portion calculate the center of the light-shielded portion as the center of an actual pointer, and use that center as angle information in coordinate calculations.
However, in the technique for detecting an angle from the peak of the light amount distribution of a light-shielded portion or from the center of the light amount distribution specified by the two ends of the light amount distribution associated with a light-shielded shadow, and calculating the pointed coordinate position based on a combination of angles detected from respective light-receiving units like the conventional light shielding method, when a plurality of coordinate positions, e.g., at least two coordinate positions are to be simultaneously input, these two input points often overlap each other to nearly line up when viewed from a given light-receiving unit.
Hence, when light-shielded shadows for the two input points overlap each other from the perspective of the light-receiving unit, these light-shielded shadows cannot be separated to detect the angles of the respective input points, thus disabling coordinate inputs.
A practical example of such case will be explained below using FIG. 22.
For example, when positions on a coordinate input region shown in FIG. 22 are respectively pointed using pointers A and B, the light amount distributions corresponding to the pointers A and B at the position of a light-receiving unit S2 in FIG. 22 are as indicated by A and B in FIG. 23B, and light-shielded shadows corresponding to two light-shielded positions of the pointers A and B are separately detected.
Note that, as reference data, the light amount distribution when no designation input is made is as shown in FIG. 23A. A valley of the light amount distribution at a position C corresponds to the light amount distribution generated due to factors such as the angle characteristics of the retroreflecting member formed around the coordinate input region, attenuation due to a distance, and the like.
On the other hand, the light amount distributions corresponding to the pointers A and B in case of a light-receiving unit S1 shown in FIG. 22 are as shown in FIG. 23C, and light-shielded shadows corresponding to the two point positions of the pointers A and B are detected to overlap each other. In the information of the light amount distribution (shielded light amount distributions) having such overlapping light-shielded shadows (shade overlapping), when A and B in FIG. 23B partially overlap each other (when a so-called partial eclipse has occurred), one end information of the light-shielded range of each pointer is obtained, as shown in FIG. 23C. For this reason, for the conventional method of calculating the position (angle) on the basis of the center or the central pixel number from information of the two ends of the light-shielded range, it is impossible to calculate the coordinate points of the pointers A and B.
Although not shown, when the shadow of the first pointer on the front side completely includes that of the second pointer farther from the light-receiving unit with respect to the target light-receiving unit (when a so-called total eclipse has occurred), the central position (angle) of the first pointer on the front side can be calculated based on the two ends of its light-shielded shadow, but information associated with the farther second pointer cannot be obtained.
Therefore, in the above prior art, the number of light-shielded shadows generated by simultaneous inputs of a plurality of pointers is detected in advance. For example, when the number of shadows detected by the second light-receiving unit is “2” and that detected by the first light-receiving unit is “1”, it is determined that the light-shielded shadows corresponding to the pointers overlap each other in the light amount distribution to be detected by the first light-receiving unit.
In such case, Patent Registration No. 2896183 adopts an arrangement which calls user's attention by generating an alarm indicating occurrence of such state, and avoids that state. In Japanese Patent Laid-Open No. 2002-055770 or 2003-303046, the first light-receiving unit must be switched to another third light-receiving unit that can detect two separated light-shielded shadows free from any overlapping, and the angles are detected using the light-receiving units (in this case, the first and third light-receiving units) that can detect these two light-shielded shadows. The aforementioned true/false determination must then be applied to the input coordinate candidates input from the respective light-receiving units to determine the two final real input coordinate positions.
In this case, since the true/false determination can be sufficiently done based on angle information of the light-receiving unit that detects shade overlapping, Japanese Patent Laid-Open No. 2003-303046 or Patent Registration No. 2896183 performs true/false determination using the angle information of this light-receiving unit that detects shade overlapping.
As described above, in the coordinate input apparatus of the light-shielding method, for example, when two-pointers are simultaneously used to input, if the light-receiving units suffer shade overlapping even merely partially, and light-shielded shadows corresponding to the two pointers are connected and cannot be separated, coordinate detection precision deteriorates due to a deviation from an actual position if the connected light-shielded shadows are considered as that from one pointer. When still higher positional precision is required, inputs are disabled, resulting a serious disadvantage upon operations.
When angle information of the light-receiving unit that detects shade overlapping is not used, and the light-receiving unit that detects shade overlapping is switched to another third light-receiving unit spaced apart by a predetermined distance from that unit so as to calculate the coordinate position, the following problem is posed.
Upon switching the light-receiving unit, discontinuity of the calculated coordinate positions occurs. In practice, since the respective light-receiving units have different characteristics, coordinate positions may become discontinuous on a region before and after the light-receiving units are switched.
The discontinuity due to the light-receiving units can be adjusted by correction to some extent if it is caused by variations of the light-receiving unit itself as a device.
However, in the aforementioned prior art, since the distance itself between the light-receiving units is used in coordinate calculations, a predetermined distance or more must be assured as that distance so as to precisely calculate the coordinate position. Furthermore, in order to allow at least one light-receiving unit to separately detect two light-shielded shadows on the coordinate input region, the light-receiving units must be arranged to be spaced apart by the predetermined distance or more. For this reason, such layout causes variations of detected light amount distributions, which are more likely to influence the discontinuity of the calculated coordinate positions upon switching the light-receiving units.
Another problem posed when angle information of the light-receiving unit that detects shade overlapping is not used, and the light-receiving unit that detects shade overlapping is switched to another third light-receiving unit spaced apart by a predetermined distance from that unit so as to calculate the coordinate position is deterioration of the coordinate detection precision resulting from the relationship between the light-receiving unit positions and the coordinate input region.
For example, as shown in FIG. 24, when coordinates are input at positions 1 and 2 of the coordinate input region as a combination of angle information of light-receiving units 1 and 2, which are arranged near the two, i.e., left and right ends of one side of the coordinate input region by normal single pointing, a given error associated with an angle of each light-receiving unit is not so expanded, and the degree of the influence on the calculated coordinate position is small.
Furthermore, in case of a plurality of simultaneous inputs, when the light-receiving unit S1 farther from the pointed position detects shade overlapping, as shown in FIG. 22, the light-receiving unit S1 that detects shade overlap is switched to the light-receiving unit S3 shown in FIG. 22, thus avoiding the problem of deterioration of the coordinate detection precision resulting from the relationship between the light-receiving unit positions and the coordinate input region as in the case of FIG. 24.
However, when the light-receiving unit S2 closer to the pointed position detects shade overlapping, as shown in FIG. 25, the light-receiving unit S2 is switched to the light-receiving unit S3, as shown in FIG. 26. However, in this case, especially, as for pointing at the position of the pointer A, the angle defined by the light-receiving unit S1, pointer A, and light-receiving unit S3, indicated by the bold lines that pass the center of the pointed position, becomes extremely small, and the influence of an error increases, as is geometrically apparent. This is more likely to considerably deteriorate the coordinate calculation precision.
Furthermore, depending on the structure and specification of a display integrated with the coordinate input device, it is often difficult to assure a space for arranging the light-receiving unit, which is selectively used upon detection of shade overlapping, at the central portion between the light-receiving units at the left and right ends of the upper or lower side of the conventional coordinate input region.
The light-receiving unit which is arranged at the central portion must have a broader detection range than those of the light-receiving units arranged at the corner portions. For this reason, in order to optically assure a field angle approximate to 180° by a single light-receiving unit, a substantial optical path length with the coordinate input region is prolonged by, e.g., a mirror arrangement, or a plurality of light-receiving units must be adopted to share the visual field range. In case of this mirror arrangement or the plurality of light-receiving units, a broader installation space around the display is required, and a so-called picture frame size increases.